Influenza virus infections remain a major global health concern despite the availability of vaccines and substantial preexisting immunity in the human population. Although current influenza vaccines or natural infection can elicit high titers of neutralizing anitbodies and virus-specific T cells, these mechanisms sometimes fail to provide protective immunity. Similarly to influenza, vaccination regimens against many other pathogens use a prime-boost strategy where preexisting immunity shapes the recall immune response. Despite the importance of these approaches for generating durable protective immunity, we do not fully understand how differing levels of preexisting immunity impacts the quantity and quality of the recall response to infection or vaccination. Furthermore, we do not have a comprehensive quantative framework to describe how preexisting humoral and cellular immunity alters the dynamics and evolution of the recall response. Our inability to make quantative predictions based on experimental evidence is a critical problem for vaccinology, as the rational design of vaccines must seek to optimize parameters in order to achieve optimal cellular immunity. We will use well-defined experimental models of influenza infection to test and refine mathematical models that will enable us to accurately predict the recall immune response in the presence of preexisting immunity. The overall goal of this proposal is to develop quantitative models that will aloow us to predict vaccine efficacy and we will do first steps toward it through three Specific Aims. In Aim 1 we will test epitope-masking model prediction that at low antigen doses the antibody responses to conserved epitopes on hemagglutinin, influenza surface protein that was shown to be a major antibody target, are not boosted. Aim 2 will test how prior CD8 T cell immunity (especially memory resident T cells in the lungs) affects recall responses to influenza. In Aim 3 we will extend the studies of Aim 1 and 2 to consider the effect of having both prior humoral and T cell immunity on the dynamics of virus and immunity following challenge with different strains of influenza viruses. For the first time we will quantify the relationship between HAI titer / preexisting CD8 T cell immunity and clearance of virus in vivo. Finally, Aim 4 will test the consistency of our verified and refined in mouse system models using humans samples banked from studies on vaccination with trivalent influenza vaccines and natural influenza infection.